India, Sept. 7 -- A single week in early September saw four of the world's leading artificial intelligence labs, Anthropic, Meta, Google and OpenAI, all release major new models in quick succession, feeding growing talk within the industry of "model fatigue" among the enterprise buyers and developers expected to keep pace with each new release.

Anthropic opened the week's activity by releasing Claude Fable 5.1 and Claude Mythos 5.1, which it described as its most advanced models for coding and knowledge work to date. Meta followed the next day with Muse Spark 1.3, while Google introduced Gemini 3.8 Flash in the same burst of activity, before OpenAI capped the week by unveiling GPT-6 Astra, which it billed as its most capable and aligned ...